Modeling nitrogen deposition with an ultra-fine mesh: application in farmland with dense river network.

Abstract

Ammonia (NH₃) emissions from agriculture significantly impact aquatic ecosystems through the deposition of ammonia nitrogen (NH_x-N). Models established on a gridded map provide a useful tool to investigate NH₃ dispersion and deposition. However, coarse model grid cells in major atmospheric chemistry models often lead to large bias in modeled N burden due to NH_x-N deposition in small household farms, which are typical in Southeastern China with dense farmland water networks. Here, we first simulated atmospheric concentrations of NH₃ and NH₄⁺ in 2020 in Changshu, a representative area with developed agriculture and dense river and farmland water networks attached to the Taihu Lake Basin in Southeast China, using the WRF-CMAQ model. To address the bias caused by coarse model grids, we employed a large leaf model to calculate dry deposition velocities for NH₃ and NH₄⁺ at a high resolution of 10 meters. This allowed for a more accurate estimation of the nitrogen load to surface waters from re-emitted farmland NH₃. The results show that coarse particulate NH₄⁺ (CNH₄⁺) exhibited the largest dry deposition velocity, followed by gaseous NH₃ and fine particulate NH₄⁺ (FNH₄⁺). The dry deposition velocities of NH₃ and NH₄⁺ were significantly influenced by land use type. Our results indicate that approximately 118.0 tons of nitrogen from ammonia reemission from farmlands entered surface water in 2020 in Changshu via atmospheric NH_x-N deposition. The wet deposition of NH₄⁺ was the largest contributor (49.8%), followed by dry deposition of gaseous NH₃ (20.9%), CNH₄⁺ (20.7%), and FNH₄⁺ (8.6%), respectively.